Survey Exploration
Juan Rocha
2016-07-15
BEST: Survey exploratory analysis
The BEST project aims to understand what are the behavioural responses of people who rely on natural resources for their livelihood, and whose resources could potentially present threshold like dynamics, aka. “regime shifts”. For this study we designed a game played with fishermen in four communities of the Colombian caribbean. Here I present a short exploration of the game data and survey.
Where?
The game was performed in four coastal communities of the Colombian Caribbean. Taganga is traditional fishing town that in the last decades has become more of a turistic hub for backpackers. Fishing is still important in the area and is mainly done in the sea, sometimes on the marine territory of Tayrona National Park. Tasajeras is a small fishing community located on the road that communicates Santa Marta and Barranquilla, two major cities. Their location favors fishing both in the sea and inside the Cienaga Grande de Santa Marta, a wetland complex of brakish water, mangroves and a very special ecosystem that depends on the exchange of salt water from the sea and fresh water from the Magdalena river. The inhabitants of Buenavista literally live on the Cienaga. Their houses stand just above water on an area of the Cienaga that is flooded all year round. They are the poorest community we visited and they depend completely on explotation of natural resources for survival. Las Flores, in contrast, is actually a neighbourhood of Barranquilla located in the outskirts of the city. They fish on the Magdalena river’s mouth with very special techniques that use wind and kites to fish on the sea. They also exploit resources in neigbouring wetlands formed by the delta.
We designed the experiment to test for differences on treatment, but also to test for difference on ecological conditions, expecting that communities who rely more strongly on the Cienaga (a closed system that has undergone regime shifts in the past) would behave differently that communities exposed more to oceanic conditions; where variability has perhaps masked/buffered regime shifts.
load('~/Documents/Projects/BEST - Beijer/BEST/mapBEST.RData')
ggmap(map) +
geom_point(aes(x=lon, y=lat, colour='Orange', size=2), data=coords, alpha=0.5,
show.legend = F) +
geom_text(aes(x=lon, y=lat), data= coords ,
label=c('Taganga', 'Tasajeras', 'Buenavista', 'Las Flores'),
size=4, colour='blue', nudge_y = 0.03) + ggtitle('Study area')
A graphical exploration of the survey
We ran a survey with each of the 256 participants of our fishing game. The 56 questions of the survey are grouped around five topics:
- the game
- fishing activities
- changes in resources and traditional ecological knowledge
- cooperation and community activities
- demographic information at household level
For each section I present a series of graphic summaries for questions whose answer was binary, ordinal or numeric. Later I elaborate on more detailed information for text data.
1. The game
Most people was not expecting the end of the game and wanted to play at least 5 more rounds. That’s good as control for end of game effect.
grid.arrange(q.num (dat = surv, key = key, q = 19),
q.num (dat = surv, key = key, q = 20),
q.num (dat = surv, key = key, q = 21),
q.num (dat = surv, key = key, q = 22),
q.num (dat = surv, key = key, q = 23) + xlab("0 = None; 1 = < 5 rounds; 2 = > 5 rounds"),
nrow=2, ncol=3)
Also an inspection of the most frequent words used when people was asked about any particular comments about the game reveals that in general people like it and see it as a learning experience.
oldpar <- par()
par (mfrow= c(1,3), family = "serif")
q.txt (dat = surv, key = key, q = 24)
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : cuidar could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pescadores could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : permite could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : conocimiento could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : aprende could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : despejar could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : comida could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : quedamos could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : agunos could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : dividieron could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pensando could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : teoricamente could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : diversix97n could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : mejora could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : diario could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : dinamico could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : opinix97n could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pescado could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : prueba could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : bastante could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : ensex96x97 could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : incentivo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : recree could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : actividad could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : reproduccix97n could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : compartir could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : explicar could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : especies could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : ninguno could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : mancha could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : acostumbrados could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : bacano could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : considera could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : ilustrativo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : ayudo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : aprendio could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : mantener could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : dejar could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : sacrificarse could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : matematico could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : interesante could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : ecosistema could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : primera could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : contientizx97 could not be fit on page. It will not be plotted.
q.txt (dat = surv, key = key, q = 28)
q.txt (dat = surv, key = key, q = 42)
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : turismo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : trabajo could not be fit on page. It will not be plotted.
par (oldpar)
## Warning in par(oldpar): graphical parameter "cin" cannot be set
## Warning in par(oldpar): graphical parameter "cra" cannot be set
## Warning in par(oldpar): graphical parameter "csi" cannot be set
## Warning in par(oldpar): graphical parameter "cxy" cannot be set
## Warning in par(oldpar): graphical parameter "din" cannot be set
## Warning in par(oldpar): graphical parameter "page" cannot be set
2. Fishing activities
Most people started fishing as teenager, however it varies depending of place. The density plot reveals such differences, fishermen in Taganga starts fishing younger, while the oldest at start are fishermen from Las Flores. Most people have been fishermen all then time since they started, and most of them have also been fishing on the same community. For less than half of our sample, people stop fishing at least a month last year (2015)
a0 <- dplyr::select(surv, id = 235, q = 25, place = locationName, treatment = treatmentName)
if (key$Data.type [25] == "binary" ) a0$q <- as.factor(a0$q)
if (key$Data.type [25] == "ordinal" ) a0$q <- as.factor(a0$q)
grid.arrange(
q.num (dat = surv, key = key, q = 25),
ggplot( data = a0, aes (x= q, fill = place)) +
geom_density(alpha = 0.4) + theme_minimal(base_size = 10, base_family = "Helvetica") +
ggtitle(key$Name.in.datasheet[25]) + xlab(key$Explanation.coding[25]),
q.num (dat = surv, key = key, q = 26),
q.num (dat = surv, key = key, q = 27),
q.num (dat = surv, key = key, q = 29),
nrow=2, ncol=3
)
## Warning: Removed 3 rows containing non-finite values (stat_density).
rm(a0)
I’d like to get a cleaner version of the fishing calendar (e.g. a heatmap), but for now the figure below summarizes when were people non-fishing last year. For all graphs 0 means that they do not stop fishing in the month, while 1 means stop fishing. The inter-monthly variability is not so high but when most people take a fishing break is in December, probably due to local festivities.
grid.arrange(
q.num (dat = surv, key = key, q = 30)+ xlab(NULL),
q.num (dat = surv, key = key, q = 31)+ xlab(NULL),
q.num (dat = surv, key = key, q = 32)+ xlab(NULL),
q.num (dat = surv, key = key, q = 33)+ xlab(NULL),
q.num (dat = surv, key = key, q = 34)+ xlab(NULL),
q.num (dat = surv, key = key, q = 35)+ xlab(NULL),
q.num (dat = surv, key = key, q = 36)+ xlab(NULL),
q.num (dat = surv, key = key, q = 37)+ xlab(NULL),
q.num (dat = surv, key = key, q = 38)+ xlab(NULL),
q.num (dat = surv, key = key, q = 39)+ xlab(NULL),
q.num (dat = surv, key = key, q = 40)+ xlab(NULL),
q.num (dat = surv, key = key, q = 41)+ xlab(NULL),
nrow=3, ncol=4
)
Most people fish 6 days a week and the most common fishing day is of 12 hours, followed by working days of 10 and 8 hours. Note that due to outliers (probably industrial fishermen) the plots for earnings and catches have their x-axis log-scaled. The minimum daily wage on an average day is Col$ 500 and the median is Col$ 30 000 which is less than US$10, while the maximum is Col 8 000 000. We believe this outlier values belong to industrial fishermen that happen to participate on our game; according to Lina Maria Saavedra from UniMagdalena, some of these values sounds realistic if the person happens to be on an industiral crew and is the owner of the boat or fishing gear. On a bad day most people don’t earn anything and the median is Col. 2000 which is less than 10 Swedish crowns. Most people have bad days once or more than once a week.
grid.arrange(
q.num (dat = surv, key = key, q = 43),
q.num (dat = surv, key = key, q = 44),
q.num (dat = surv, key = key, q = 45) + scale_x_log10() + geom_density(alpha = 0.4) ,
q.num (dat = surv, key = key, q = 46) + scale_x_log10() + geom_density(alpha = 0.4) ,
q.num (dat = surv, key = key, q = 47),
q.num (dat = surv, key = key, q = 48) + scale_x_log10() + geom_density(alpha = 0.4) ,
q.num (dat = surv, key = key, q = 49) + scale_x_log10() + geom_density(alpha = 0.4) ,
q.num (dat = surv, key = key, q = 50) + scale_x_log10() + geom_density(alpha = 0.4) ,
q.num (dat = surv, key = key, q = 51) + ggtitle('16f. How often do you have a bad day?') + xlab("1/yr = 1, 1/month = 2, 1/week = 3, >1/week = 4"),
nrow=3, ncol=3
)
Most people fish in company of other fishermen, but almost a fifth of our surveyed fishermen fish alone. Crew size varies across places, very few people fish with crews > 12, and they are from Taganga or Las Flores, the fishing towns where open sea fishing with more industrial styles are possible. In fact, the maximum crew size for Tasajera is 5 and in Buenavista is 3 or 4. Most people fish with the same crew, but most of the time the decision on when or where to go fishing is made by other person than the intervieweed. The most typical boat is bote which implies a small size vessel without engine or with a small power engine. Only few people (<20) fish on lanchas which are bigger boats with more powerful engines, most of them are located in Las Flores where wind and currents are strong due to the Magdalena river delta. Similarly, around 60% of fishermen are not the captain neither own the boat they fish with. They don’t fish on the same places either. Half of the people own their fishing art; ownership is most common in Buenavista and less common in Taganga.
grid.arrange(
q.num (dat = surv, key = key, q = 52),
q.num (dat = surv, key = key, q = 53),
q.num (dat = surv, key = key, q = 54) ,
q.num (dat = surv, key = key, q = 55) ,
q.num (dat = surv, key = key, q = 56),
q.num (dat = surv, key = key, q = 58),
q.num (dat = surv, key = key, q = 59),
q.num (dat = surv, key = key, q = 60),
q.num (dat = surv, key = key, q = 61),
nrow=3, ncol=3
)
The graphs below summarize the most important species and some other textual answers from previous sections of the survey. The first row shows the most common answers for the questions about their secondary activity when they are not fishing (left), who decides where and when to fish (center) and the most common fishing ground. Rows 2-5 show on the left the most important species, the fishing art used (center) and transportation means to reach the fishing grounds (right).
# quartz(width=5, height = 5, family = "serif", pointsize = 6)
oldpar <- par()
par (mfrow= c(5,3), family = "serif")
for (i in 1:15){
x <- c(42, 57,62:65,68, 69,70,73:75,78:80)
q.txt (dat = surv, key = key, q = x[i])
}
par (oldpar)
The graphs below are a bit weird to interpret. We asked the fishermen what were the most important species they fish, their names appear above. Afterwards we asked them to rank them from most important (1) to less important (4), and importance were divided by biomass in kg or by money earned. So the results only make sense when looking at individual interviews, at the aggregated level they are confusing.
grid.arrange(
q.num (dat = surv, key = key, q = 66) + xlab(NULL) + ggtitle("Importance spp1 Kg") ,
q.num (dat = surv, key = key, q = 67) + xlab(NULL) + ggtitle("Importance spp1 $$"),
q.num (dat = surv, key = key, q = 71)+ xlab(NULL) + ggtitle("Importance spp2 Kg"),
q.num (dat = surv, key = key, q = 72)+ xlab(NULL) + ggtitle("Importance spp2 $$"),
q.num (dat = surv, key = key, q = 76) + xlab(NULL)+ ggtitle("Importance spp3 Kg"),
q.num (dat = surv, key = key, q = 77) + xlab(NULL)+ ggtitle("Importance spp3 $$"),
q.num (dat = surv, key = key, q = 81) + xlab(NULL)+ ggtitle("Importance spp4 Kg"),
q.num (dat = surv, key = key, q = 82) + xlab(NULL)+ ggtitle("Importance spp4 $$"),
nrow = 4, ncol= 2
)
Where does the food goes? Most fishermen take some fish home but they usually sell more than half and ocassionally give some away. When asked if they imagine themeselves being fishermen in 10 years most people think they will be. Question 28 is horribly coded, it should be between 0:4 but has values >5 up to 10! There is also typing errors in Tasajera where many values were included in answers that should have been unique.
grid.arrange(
q.num (dat = surv, key = key, q = 83),
q.num (dat = surv, key = key, q = 84),
q.num (dat = surv, key = key, q = 85),
q.num (dat = surv, key = key, q = 87),
q.num (dat = surv, key = key, q = 88),
q.num (dat = surv, key = key, q = 90),
nrow=2, ncol=3
)
The following word clouds show who is the main person who buys the fish (left). For fisherment who answered that they don’t see themeselves fishing in the future, we asked why not? their most common words used are shown in the middle; and for fishermen who aswered that they don’t see their kids fishing in the future, their most common words used are plotted in the right.
oldpar <- par()
par (mfrow= c(1,3), family = "serif")
q.txt (dat = surv, key = key, q = 86)
q.txt (dat = surv, key = key, q = 89)
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : futuro could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : produccion could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : disminucix97n could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : maquinaria could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : construcciones could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : tener could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : vueltas could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pequex96as could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : dedicarme could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : perdida could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : cansado could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : consigue could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : problemas could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : portuarias could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : vigilante could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estudiando could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : trabajo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : profundas could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : subsistencia could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : viene could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : personal could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estudia could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : profesion could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : quieres could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : agotando could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pescarx87 could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : perdiendo could not be fit on page. It will not be plotted.
q.txt (dat = surv, key = key, q = 91)
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : quiere could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : disminux92do could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estx87 could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estudiaran could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : tambix8en could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : desgastante could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : disminuye could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estudie could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pocoy could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : perspectiva could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : buenavista could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : hembras could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pueden could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : capacito could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : prepare could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : trabajen could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : llevado could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estan could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : mas could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : riesgo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : studiando could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : mejor could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : cansada could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : gusta could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : espero could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : actualidad could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pescan could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pescado could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estudien could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : encuentren could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : puedan could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : acabarx87 could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : superen could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : marta could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : verlos could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : peces could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pasx8e could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : asx92 could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : gustaria could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : adelante could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : trabajopero could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : profesional could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : comerciante could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : inestable could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : deseo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : situaciones could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : condiciones could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estudio could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : bienestar could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : proyecciones could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : ejemplo could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : gustarx92a could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : murio could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : mejores could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : sigan could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : futuro could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : actividad could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : dufre could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : quisera could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : hijos could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : agotando could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : pescadores could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : oficio could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : preparados could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : estx87n could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : varones could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : darle could not be fit on page. It will not be plotted.
## Warning in wordcloud(dat, scale = c(3, 0.5), min.freq = 0.1, random.order =
## T, : riesgoso could not be fit on page. It will not be plotted.
par (oldpar)
## Warning in par(oldpar): graphical parameter "cin" cannot be set
## Warning in par(oldpar): graphical parameter "cra" cannot be set
## Warning in par(oldpar): graphical parameter "csi" cannot be set
## Warning in par(oldpar): graphical parameter "cxy" cannot be set
## Warning in par(oldpar): graphical parameter "din" cannot be set
## Warning in par(oldpar): graphical parameter "page" cannot be set
3. Changes in resources & TEK
When we asked about species dynamics people reported that there has been speces that they don’t fish as much as before. More than 150 respondants report this changes as dramatic. On the left you see the word cloud for the most frequent words used when asked what has changed?, in the center which species have changed? and how? (left). The following rows and columns correspond to the answers in question 35 which is a table. Each respondant was asked what happened? (left), what do they think were the causes (center), and how were they affected? (left) for a maximum of 4 of this abrupt events or drastic changes on resouce availability. Accordingly, for each of the 4 changes mentioned, we also asked when did it happen? and how long has been the resource missing?. The later questions are numerical and summarized below the word clouds.
oldpar <- par()
par (mfrow= c(5,3), family = "serif")
for (i in 1:15){
x <- c(93,95,97,118,100:102,105:107,110:112,115,116)
q.txt (dat = surv, key = key, q = x[i])
}
par (oldpar)
grid.arrange(
q.num (dat = surv, key = key, q = 92) + xlab("Yes=1, No=0"),
q.num (dat = surv, key = key, q = 94) + xlab("Yes=1, No=0"),
q.num (dat = surv, key = key, q = 96) + xlab("Yes=1, No=0"),
# q.num (dat = surv, key = key, q = 97),
q.num (dat = surv, key = key, q = 98),
q.num (dat = surv, key = key, q = 99),
q.num (dat = surv, key = key, q = 103),
q.num (dat = surv, key = key, q = 104),
q.num (dat = surv, key = key, q = 108),
q.num (dat = surv, key = key, q = 109),
q.num (dat = surv, key = key, q = 113),
q.num (dat = surv, key = key, q = 114),
nrow=3, ncol=4
)
Responses on abrupt events on fishing: In question 35 people were asked if they have experienced an abrupt change (above graph). Questions 36:40 were only asked to people who did experience abrupt changes, so N != 256. People were asked if the abrupt event described on table/question 35 have affected what they fish. Most people answer positively, nevertheless about 40 people say ‘no’. Since the occurrance of the event, most people spend more time fishing, less than 20 spends the same or less time respectively. Just above 60 people changed their fishing area while just below 50 didn’t. Questions 37 to 39 explore the hypothetical case of drastic events in the future. Most people expected abrupt changes in the future but they strategies on how they would deal with it varies: In case of an abrupt event fewer people would continue fishing, most of them won’t increase effort, but they won’t reduce their fishing time neither, less than 40 would consider changing fishing grounds or fishing arts, and most of them wont stop fishing.
grid.arrange(
q.num (dat = surv, key = key, q = 117),
q.num (dat = surv, key = key, q = 119) + ggtitle("more or less time fishing?"),
q.num (dat = surv, key = key, q = 120),
q.num (dat = surv, key = key, q = 121) + ggtitle("37. Expect changes?"),
#q.num (dat = surv, key = key, q = 123),
q.num (dat = surv, key = key, q = 124),
q.num (dat = surv, key = key, q = 126),
q.num (dat = surv, key = key, q = 128),
q.num (dat = surv, key = key, q = 130),
q.num (dat = surv, key = key, q = 132),
q.num (dat = surv, key = key, q = 134),
# q.num (dat = surv, key = key, q = 137),
nrow=2, ncol=5
)
Question 38 explores what they think would be the cause of the hypotehtical abrupt event, being climate change, overfishing and pollution among the expected causes. The rest of the word clouds are the answer of open questions that follows up the multiple selection described above. So for example, if the respondant said in question 39.2 that she or he will continue fishing in case of the drastic event, the world cloud below for 39.2 explores the answers to the follow up question why?.
oldpar <- par()
par (mfrow= c(3,3), family = "serif")
for (i in 1:8){
x <- c(122,125,127,129,131,133,135,136) # 144 not ploted
q.txt (dat = surv, key = key, q = x[i])
}
par (oldpar)
4. Cooperation and community
Most fishermen shared their fishing arts, but only ~100 belong to a fishing coop. The community with stronger coop presence is Las Flores (note however that our local contact was the coop leader!) while the community with less cooperative participation is Buenavista. People participating in coops usually meet once a month followed by twice a year. Surprisingly, most participants self-report to be treasurers (but there is only one treasurer per coop!?). The graphs about the reasons to belong to a coop doesn’t make much sense. Future versions should compare (bars) for positive answers, here I’m not sure if negative was coded as default or that they did not perceive the benefit. For example, for ‘better prices’ only 20 people report that benefit while ~80 said ‘no’ to that question.
grid.arrange(
q.num (dat = surv, key = key, q = 137) + xlab(NULL),
q.num (dat = surv, key = key, q = 139)+ xlab(NULL),
q.num (dat = surv, key = key, q = 141) + xlab("1/year = 1, 2/year = 2, 1/month = 3, 1/week = 4"),
q.num (dat = surv, key = key, q = 142),
q.num (dat = surv, key = key, q = 143) + xlab ("0 = representant, 1= President, 2= secretary,\n 3= treasurer, 4= active member, 5= other"),
#q.num (dat = surv, key = key, q = 146),
q.num (dat = surv, key = key, q = 147)+ xlab(NULL) ,
q.num (dat = surv, key = key, q = 148) + xlab(NULL),
q.num (dat = surv, key = key, q = 149)+ xlab(NULL),
q.num (dat = surv, key = key, q = 150)+ xlab(NULL),
q.num (dat = surv, key = key, q = 151)+ xlab(NULL),
q.num (dat = surv, key = key, q = 152)+ xlab(NULL),
q.num (dat = surv, key = key, q = 153)+ xlab(NULL),
q.num (dat = surv, key = key, q = 154)+ xlab(NULL),
nrow=4, ncol=4
)
The word clouds below correspond to the questions who do you ask for help when needed?, which coop do you belong to?, if the respondant said that he/she has another role in question 42.4, here we follow up with which other role?, and the last one summarizes the benefits of belonging to a coop.
oldpar <- par()
par (mfrow= c(1,4), family = "serif")
for (i in 1:4){
x <- c(138, 140,144,145 )
q.txt (dat = surv, key = key, q = x[i])
}
par (oldpar)
5. Demographics
Almost all our game participants were male fishermen, most of them married or in co-habitation with partner. The average age for the full sample is 42.7351779 and the median is 42. Our sample is younger in Taganga and older in Las Flores. Most people finished elementary school, over 25 respondents didn’t have any formal education, and less than 20 had university degrees almost all of them from Taganga (There is an error with 0 - datapoint in Tasajeras). Most respondants were born on the community but ~70 of them came from somewhere else (error data point with 10). About 40 people have been moving around while <200 have always live in the same community.
grid.arrange(
q.num (dat = surv, key = key, q = 155),
q.num (dat = surv, key = key, q = 156) + xlab ("1= single, 2= married, 3= co-habitation, 4= divorce/widowed"),
q.num (dat = surv, key = key, q = 157),
q.num (dat = surv, key = key, q = 158) + xlab ("1= none, 2= elementary, 3=high school, 4= university"),
q.num (dat = surv, key = key, q = 159),
q.num (dat = surv, key = key, q = 160),
q.num (dat = surv, key = key, q = 162),
q.num (dat = surv, key = key, q = 163),
q.num (dat = surv, key = key, q = 164),
q.num (dat = surv, key = key, q = 165),
q.num (dat = surv, key = key, q = 232) + xlab (NULL),
nrow=3, ncol=4
)
When asking if they have suffered forced displacement few respondents (n=5) report forced displacement from 6 months up to 7 years. Question 52 was designed to assess sense of place. Each participant was asked to agree or disagree on a scale from 1 to 4 (1=strongly agree, 2=agree, 3=disagree, 4 strongly disagree) with statements that were read outloud for them. The figure below summarize their answers per location.
grid.arrange(
q.num (dat = surv, key = key, q = 166) + xlab (NULL),
q.num (dat = surv, key = key, q = 167) + xlab (NULL),
q.num (dat = surv, key = key, q = 168) + xlab (NULL),
q.num (dat = surv, key = key, q = 169) + xlab (NULL),
q.num (dat = surv, key = key, q = 170) + xlab (NULL),
q.num (dat = surv, key = key, q = 171) + xlab (NULL),
q.num (dat = surv, key = key, q = 172) + xlab (NULL),
q.num (dat = surv, key = key, q = 173) + xlab (NULL),
q.num (dat = surv, key = key, q = 174) + xlab (NULL),
nrow=3, ncol=3
)
The only information that has not been ploted from the survey is the table 53 where we ask about the household composition and income.